Background With the socioeconomic development in China, the prevalence of overweight/obesity has risen significantly, becoming a major public health issue. Currently, commonly used obesity indicators, such as BMI and waist circumference (WC) cannot differentiate between subcutaneous fat and visceral fat, while visceral fat is closely associated to chronic diseases. Therefore, exploring the correlation between novel anthropometric indices and stroke risk holds significant importance.
Objective To explore the correlation of different anthropometric indices with the incidence of stroke, and their predictive capacity for the risk of stroke, aiming to provide evidence for community-based chronic disease health management and cardiovascular/cerebrovascular disease surveillance.
Methods Based on the Pudong New Area Chronic Disease and Risk Factors Surveillance Cohort Study Project, a nested case-control study was performed involving participants enrolled in the 2016 and 2019 field surveys. They were followed up until December 31, 2023. Individuals who developed stroke during follow-up were assigned to the case group, and those without developing stroke were served as controls. Data on demographics, medical history, family history, and major risk factors were collected using standardized epidemiological questionnaires. Physical and laboratory examination indicators were recorded. Logistic regression and restricted cubic spline (RCS) regression models were applied to analyze the correlation between anthropometric indices and stroke. Receiver operating characteristic (ROC) curves were used to evaluate the predictive performance, with pairwise comparisons conducted via DeLong's test.
Results Among the 15 440 study subjects included in the analysis, a total of 930 had strokes. For every one-unit increase in BMI, WC, body roundness index (BRI), and Chinese visceral adiposity index (CVAI), the risk of stroke increased by 3.8% (OR=1.038, 95%CI=1.017-1.058), 1.2% (OR=1.012, 95%CI=1.004-1.020), 10.6% (OR=1.106, 95%CI=1.042-1.174), and 0.5% (OR=1.005, 95%CI=1.003-1.007), respectively (P<0.05). The RCS regression model showed a linear dose-response correlation of BMI, WC, and BRI with the risk of stroke (Ptotal<0.05, Pnon-linear >0.05), and a non-linear dose-response relationship between CVAI and the risk of stroke (Ptotal<0.001, Pnon-linear=0.009). The ROC curve results indicated that the predictive ability of CVAI for the risk of stroke (area under the curve=0.66) was better than that of BMI (Z=-12.713, P<0.001), WC (Z=-13.512, P<0.001), and BRI (Z=-8.696, P<0.001).
Conclusion BMI, WC, BRI, and CVAI are correlated with the risk of stroke. CVAI is superior to BMI, WC, and BRI in predicting the risk of stroke and can be used as an applicable indicator for the risk of stroke. These findings underscore the need for comprehensive chronic disease management in community health programs, emphasizing weight control and the impact of visceral obesity.